How to not get bamboozled by the question of 'what exists'

One of the more viral memes from Yuval Noah Harari’s Sapiens is a passage early on in the book that describes how “fictions” run our world. He talks about the car company Peugeot, and how when you look closely, it doesn’t really “exist.” Harari says, if Peugeot exists, where is it? You might say “Peugeot is all the people working for it,” but the employees can change entirely over the course of a few decades, and it’s still the same company. You might say “the factories are the company,” but individual factories are opened and closed all the time. You could go on and on in this way, and Harari’s conclusion from this is that Peugeot doesn’t “really exist,” that it’s a fiction much like Harry Potter. And he says the same thing is true about countries, and even entities like the US dollar. In Harari’s view, what makes humans different from other creatures is our ability to collectively believe fictions like this.

Depending on how much you’ve thought about this, Harari’s point that will sound either (1) profoundly revelatory, (2) trivial, or (3) insightful but slightly misleading. Let me explain why I take the last perspective, and how I think about the broader question of “what really exists.” (Spoiler: I don’t think it’s a good question.)

First let’s consider the insight in Harari’s statements. Yes, in a strict sense it’s true that you can’t find Peugeot in any individual person, place, or object. But there are two things to add here: (1) the fact that it’s a “fiction” in this sense says nothing about its utility as a concept; (2) this “deconstructive” line of questioning can be taken towards literally any other concept, and at some point everything breaks down into a “fiction.”

So, when Harari says “the US dollar is a fiction,” and that, for example, “people are fighting wars over a fictional entity”, he is making humans seem much more irrational than they really are. Ford and the US dollar are not “fictions” in the same way that Harry Potter is. Because these “fictions” are very deeply embedded into the real world. Ford is an actual company that you can sue, it’s a company that thousands of people work for and get their biweekly paychecks from. The US dollar is a thing that you can actually use to obtain food or get packages delivered or have people complete tasks for you, it’s something that keeps the entire economy functioning. The US dollar and Ford are fictions that people very strongly believe in, and as a practical matter that makes them a very different kind of thing than Harry Popper or Spongebob Squarepants. To be fair, Harari does acknowledge that Ford and the US dollar are “intersubjective” entities—as in, they get their “realness” from the fact that people believe in them—but in general when people talk about this idea, they don’t emphasize enough the distinction between fictions like the dollar and fictions that are more commonly recognized as “made up” like Harry Potter.

So on one hand, Harari’s point is true, but just not that useful, because calling the dollar a fiction doesn’t (and shouldn’t) change our behavior towards it. ¹

Now let’s go in the other direction and explain why this whole point about “fictions” is not just true of the dollar, but is true of literally everything else. When Harari presents his view, he takes for granted a “commonsense ontology” in which everyday objects like trees and ice cream cones very obviously “exist.” This is a reasonable assumption, but if you’re going to take the skeptical attitude of asking “does it really exist?” towards the US dollar and Ford, you should be willing to do the same for everyday objects. And indeed, once you take this attitude, it becomes clear that even everyday objects don’t exist in the way we think they do.

There are a few ways to see why.

First of all, naive realism—the view that our perceptual senses are a direct window into the objective external world—is definitively false. The simplest counterexample is visual illusions: things can look a certain way but actually be a different way (one box can seem to be bigger than another, but in actuality they are the same size). How far does this tendency to succumb to illusions go? How badly can we be deceived by our senses? It likely goes all the way down: the cognitive scientist Donald Hoffman argues that evolution specifically optimizes against seeing the world as it really is. Our perception was fine-tuned for fitness, which is very different from truth:

Recent studies of perceptual evolution, using evolutionary games and genetic algorithms, reveal that natural selection often drives true perceptions to extinction when they compete with perceptions tuned to fitness rather than truth: Perception guides adaptive behavior; it does not estimate a preexisting physical truth. Moreover, shifting from evolutionary biology to quantum physics, there is reason to disbelieve in preexisting physical truths: Certain interpretations of quantum theory deny that dynamical properties of physical objects have definite values when unobserved. In some of these interpretations the observer is fundamental, and wave functions are compendia of subjective probabilities, not preexisting elements of physical reality. These two considerations, from evolutionary biology and quantum physics, suggest that current models of object perception require fundamental reformulation.

There are various lines of research that vaguely converge onto this same thesis, that the real structure of the world is not at all like the everyday object-centered ontology that we impose on it. Karl Friston makes the argument that our perception of the world is fundamentally a process of active inference – we construct our perception of the world rather than passively receiving it. Neuroscientist Anil Seth makes a similar point when he says that our perception is a “controlled hallucination.” Brian Cantwell Smith, a philosopher who has spent decades studying the foundations of computation, also argues objects are not as cut-and-dry as we expect them to be.

And finally, even putting the sophisticated neuroscience and physics arguments aside, there’s a more straightforward way to appreciate that ordinary objects and categories are not what we think they are, which is by just inquiring about their boundaries. David Chapman talks about this in Objects, objectively, quoting Richard Feynman:

What is an object? Philosophers are always saying, “Well, just take a chair for example.” The moment they say that, you know that they do not know what they are talking about any more. The atoms are evaporating from it from time to time—not many atoms, but a few—dirt falls on it and gets dissolved in the paint; so to define a chair precisely, to say exactly which atoms are chair, and which atoms are air, or which atoms are dirt, or which atoms are paint that belongs to the chair is impossible. So the mass of a chair can be defined only approximately.

There are not any single, left-alone objects in the world. If we are not too precise we may idealize the chair as a definite thing. One may prefer a mathematical definition; but mathematical definitions can never work in the real world.

And there’s also Doug Hofstadter’s book Surfaces and Essences, in which he spends several hundred pages describing how the ways that we split up the world into categories and concepts is not at all an inherent property of the world itself, and depends largely on cultural context. Different languages chunk up the world in different ways, e.g. even with the basic concept of “going somewhere”:

In English, we can say without any sense of oddness: “Sometimes I go to work by car, and other times on foot.” In German or Russian, however, these two forms of locomotion call for different verbs. When one takes a vehicle to arrive at one’s destination, then the verb “fahren” is used in German, whereas when one goes somewhere on foot, then the verb “gehen” is used. In Russian it’s trickier yet, because not only is there a distinction between going in a vehicle and going on foot, but also the choice of verb depends on whether this kind of motion is undertaken frequently or just one time. Thus a completely innocuous-seeming verb in English breaks up into several different verbs in Russian. In other words, what to English speakers seems to be a monolithic concept splits into four distinct concepts to Russian speakers.

The point is: Ford is a “fiction” to some extent, but then again all of our concepts are fictions to one degree or another, and some fictions are more widely used for practical purposes than others.

It goes without saying that none of the above is an argument for epistemic nihilism, the view that we know absolutely nothing about reality. We actually know quite a lot about the world and are able to make progress in garnering useful knowledge about it. But there is this inescapable layer of fictionalizing that takes place, where our perceptions and concepts impose onto the world a bunch of neatly ordered (and useful!) categories that make the world easier to work with, but which in the end are not inherent to the world itself, and instead are the byproduct of our interaction with the world. ²

Finally, this brings us to the most broad question of “what really exists”, to which my current general attitude is, “stop asking the question in such a broad way.” People love to come to a universal answer to the question, like “what really exists is atoms/spacetime/energy and the laws of physics, everything else is made up” (materialism? reductionism?), or “what really exists is our thoughts and perceptions, everything else is made up” (solipsism?), or “what really exists is all those things that are assumed to exist in our best explanations”, which is something the critical rationalists assert, and which I think is a nice-sounding and mostly useless criterion of existence.³

We don’t need an answer to “what really exists” in general—we only ever need to ask the question in specific contexts. Does Covid-19 exist or not? That’s a very practical question that people have disagreed about, and we can validate its existence with specific tools like microscopes, various molecular tests, and our general observation that people who test a certain way display predictable symptoms. Other people will ask whether “the soul” exists, and very few of the tools we use for the Covid question are relevant for this one; in this case we can invalidate its existence by appealing to the conceptual incoherence of “the soul” as a concept (although admittedly this question is a trickier one to settle because the concept itself is so ill-defined).

The point is, when we ask “what exists,” in 99% of cases what we’re really asking is “should I be doing anything differently from what I’m doing?” And we can answer that question on a case-by-case basis. The only time you can’t do this is if a philosopher is asking the question. In that case, they’re not really asking “what I should be doing differently”—instead they’re asking “what actually really exists, like in actual reality, independent of all our made-up stories?” While it’s definitely interesting to ask the question in this particular way, given that we don’t have a good answer to it after a few thousand years of questioning, I won’t hold my breath waiting for an answer.